针对家庭服务机器人工作的复杂动态环境,结合智能空间和机器人二者优势,提出了一种融合环境及其中目标资源分布的家庭智能空间全息地图构建方法。基于Rao-Blackwellized粒子滤波思想,根据建图初期机器人及其感知的全局空间特征点位置完成智能空间传感器网络节点的位姿参数标定,利用已标定传感器网络的观测值和机器人控制量进行机器人位姿估计,同时根据机器人位姿及其观测值进行全息建图。针对建图过程中及建图后的环境和目标变动,设计了基于智能空间监测的全息地图动态更新策略。实验表明,智能空间传感器网络有效地解决了建图中的动态数据关联问题,由其辅助的机器人定位和建图精度明显提高。
To deal with the complex and dynamic home environment where the service robot works,a novel concept of holographic environment map is proposed in this paper,which describes the distribution of environment and other objects or resources.The advantages of both intelligent space and service robot are combined to construct the holographic environment map.On the basis of Rao-Blackwellized particle filtering theory,both the localizations of the robot and the feature points located by the robot are utilized to calibrate the pose parameters of each camera node of the intelligent space camera network.At the same time,both the observations of the calibrated intelligent space camera network and the controls of the robot are utilized to estimate the robot localization,and both the observations and localization of the robot are utilized to build the holographic environment map.In order to deal with the movement of objects within or after mapping,the dynamic updating strategies of the holographic environment map under the monitoring of the intelligent space camera network are designed.The experimental results show that the intelligent space camera network could resolve in an efficient manner the data association of Simultaneous Localization and Mapping(SLAM) in the dynamic environment,and with the assistance of the intelligent space,the precision of both the robot localization and holographic environment mapping are improved significantly.